How to save 2D numpy array column wise instead of row wise in python?

Question

I have this array:

[[1,2,3],[4,5],[7,8,9]]

I tried to use pandas.to_csv but it gives me

1 2 3
4 5 nan
7 8 9

but I want

1 4 7
2 5 8 
3 nan 9

Well my array is dynamic, it is not of something like 3 x 3 like first row has 3 elements, second row has 7 elements and so on... — Mayank Bansal, Dec 10 '16 at 08:39
It'd be much easier to help you if you'd give us test input and desired output. As a quick guess I'd say `df.transpose().to_csv()` should be enough — Roman Pekar, Dec 10 '16 at 09:02
I don't think this is a duplicate, the question is about dealing with rows of varying lengths — SheepPerplexed, Dec 10 '16 at 09:19

score 0 · Answer 1 · answered Dec 10 '16 at 09:15

0

I would first ensure all rows have the same length:

a = [[], [1], [1,2], [1,2,3]]
max_row_length = max([len(r) for r in a])
a = [r + [None]*(max_row_length - len(r)) for r in a]

and then I would use numpy:

np.array(a).T

answered Dec 10 '16 at 09:15

SheepPerplexed

1 Answers1